STABILITY OF THE AIR FORCE KC-1B MAPPING CAMERAS 
Clarice L. Norton 
Hill Air Force Base, Utah, 
U.S.A. 
ABSTRACT 
Air Force aerial cameras must be designed and manufactured to meet demanding military conditions 
without scarifice of performance. Such requirements are specified in purchase contracts and proved by 
acceptance and environmental tests on the prototype. One of the most critical evaluations of 
stability is obtained from the geometric data of mapping cameras. On repeated calibrations, this data 
should show negligible change. An evaluation of calibrations performed on a number of cameras over a 
span of two or more decades shows that the KC-1B has reliably fulfilled the geometric requirements of 
stability. Data is discussed. 
INTRODUCTION 
The definition of the stability of a mapping camera differs from that of a reconnaissance camera. 
They both need good image quality but the mapping camera requires stable geometry - the inner orien- 
tation elements - which are considered to be unchanging. 
In a study for Commission I of the International Society of Photogrammetry and Remote Sensing, 
Hakkarainen of Finland has raised the question of stability. But how is stability defined for an 
aerial camera which is used for several decades by many operators; is subject to field maintenance 
and depot repair; and experiences many environments? What data from an Air Force mapping camera 
should be repeatable, within what limits? Should the calibrated focal length and the distortion be 
unchanging year after year? What about the fiducials which mark the Fiducial Center, the Principal 
Point of Autocollimation, and the Point of Symmetry. If they differed for repeated calibrations, how 
much would they change and why? ' 
KC-1B CALIBRATIONS 
The author believed that some of the answers to these particular questions could be in the files of 
the Calibration Laboratory at Hill Air Force Base, where records of fourteen cameras having two or more 
calibrations were located. These had a total of thirty-three calibrations using VF spectroscopic | 
plates for recording data. Three of the cameras had also been calibrated using film in the operating 
magazine. Figure 1 shows the data obtained for each camera in a summary form. Note that the column 
labeled envelope, following the average distortion column, contains all the asymmetric distortion 
values for four diagonals of the format. It is the total width of the envelope, in micrometers, | 
referenced to the average distortion of the individual calibration. This data contains various measures 
of the geometric stability of each camera. | 
CAMERA 62-045 
Data from camera 62-045 calibrations are shown in Figures 2 and 3. The average distortion and DO dt 
distortion envelope are shown in each column for the referenced calibration, with a slant line between PEL 
the values (in the first column 43/2 means the average distortion at 7.5? is +3 micrometers and the 
width of the envelope is 2 micrometers at that hole). For this camera there were four calibrations | 
within sixteen years plus one calibration using film in the fully operational camera. In comparing the EO 
calibrated average distortions and envelope widths, there are only small differences: Two to four | 
micrometers. What does differ is the position of the Principal Point of Autocollimation (PPA) and the 
Point of Symmetry (PS). Assuming a stable fiducial center, (there could be some doubt) then these two 
points Je within a circle of eleven micrometer radius when all calibrations are evaluated. Is this 
stability? 
If the years of handling and the repairs that may affect plate tipping and focal distance are con- 
sidered, the differences seem small. What should be questioned is how these small differences affect 
the precision and accuracy of mapping. That answer lies with our research photogrammetrists and 
cartographers. 
KC-1B DISTORTION CURVE 
  
. It would be helpful if the distortion curves for all lenses of the Planigon design followed a model. 
Figures 4 and 5 show that this is not the case. Figure 5 shows the graph of two very different 
distortion curves. The widest distortion envelope will not contain them both. What has been done is to 
determine each envelope with respect to its own average curve and from the individual envelopes obtain 
a maximum. The last two columns of Figure 4 show maximum envelopes taken from 36 calibrations of 13 
Cameras, 3 calibrations being on film in the operating magazine. For any average radial distertion 
curve, such as curve B in Figure 5, no value would lie outside the envelope shown. 
Point of Symmetry (PS) Referenced to the Principal Point of Autocollimation (PPA). 
84